President Donald Trump unveiled a $500 billion private-sector initiative on Tuesday aimed at transforming AI infrastructure in the US. The joint venture, called Stargate, brings together OpenAI, SoftBank, and Oracle to build 20 massive data centres and create over 100,000 jobs. Backers have committed $100 billion for immediate deployment, with the remainder spread over the next four years.
The announcement, made at the White House with SoftBank CEO Masayoshi Son, OpenAI CEO Sam Altman, and Oracle Chairman Larry Ellison in attendance, underscores America’s push to lead in AI development. Ellison revealed that the first data centres, each half a million square feet, are already under construction in Texas. These facilities aim to power advanced AI applications, including analysing electronic health records to assist doctors.
Trump attributed the project’s launch to his leadership, with executives expressing their support. “We wouldn’t have decided to do this unless you won,” Son said. However, the ambitious project arrives amid concerns over the rising energy demands of AI data centres. Trump promised to simplify energy production for these facilities, even as experts warn of potential power shortfalls across the country in the coming decade.
The announcement comes against a backdrop of surging AI investments since OpenAI’s release of ChatGPT in 2022, which sparked widespread adoption of AI across industries. Oracle and other tech stocks, including Nvidia and Dell, climbed on the news, reflecting market enthusiasm for the Stargate project.
ByteDance, the company behind TikTok, is reportedly planning a substantial $12 billion investment in AI infrastructure by 2025. According to the Financial Times, the funds will go towards acquiring advanced AI chips and enhancing model training capabilities, both domestically and abroad. A spokesperson for ByteDance refuted the accuracy of the report, calling the claims incorrect.
The company intends to allocate 40 billion yuan ($5.5 billion) towards purchasing AI chips in China, while an additional $6.8 billion will be spent overseas. Domestic semiconductor orders would largely go to Chinese suppliers, including Huawei and Cambricon, with the remainder focused on Nvidia chips modified to comply with US export restrictions.
China’s government has encouraged tech firms to source a significant percentage of their chips from local manufacturers. Meanwhile, ByteDance continues to navigate US scrutiny, with its popular app TikTok facing political pressure to be sold.
The news comes amid a broader global race for AI dominance, where investment in cutting-edge technology remains pivotal for competitive advantage.
President Donald Trump’s executive order delaying the enforcement of a US TikTok ban has created new legal uncertainties for the platform and its service providers, including Google and Apple. Signed on Monday, the order pauses for 75 days a law requiring TikTok’s Chinese parent company, ByteDance, to divest the app over national security concerns.
While the order directs the Justice Department to halt enforcement and assures app distributors of no liability during the review period, legal experts warn that the promise offers little protection. Courts do not consider executive orders binding, and Trump could alter or selectively enforce the policy at any time, potentially exposing companies to massive penalties.
The ban, passed by Congress and upheld by the Supreme Court days before Trump’s order, imposes steep fines of $5,000 per user for violations, making compliance a high-stakes gamble for service providers. Critics argue that the legal ambiguity could also open companies to shareholder lawsuits if they ignore the ban based solely on Trump’s directive.
Trump’s move has reignited tensions between the White House and lawmakers, who overwhelmingly supported the ban over fears of Chinese influence. The coming weeks may bring further legal battles and political manoeuvring as the future of TikTok in the US hangs in the balance.
The Japan Bank for International Cooperation (JBIC) has pledged up to €800 million to support the expansion of Germany’s 5G infrastructure, part of an effort to reduce reliance on Chinese technology. The project, which includes contributions from private banks across Europe and Japan, aims to build a secure and advanced telecom system for Germany.
The funding will support United Internet AG, a German telecom company, in adopting Open Radio Access Network (Open RAN) technology. This system allows seamless integration of equipment from multiple suppliers, reducing the risks of over-dependence on a single provider. A significant portion of the software involved is developed by Rakuten Group Inc., a Japanese tech firm.
Germany has relied heavily on Chinese manufacturers for 5G infrastructure, with 59% of its network sourced from Huawei and ZTE in 2022. This new initiative reflects Germany’s ambition to phase out Chinese components by 2029 and strengthen national security. JBIC’s €300 million contribution represents the largest share of the funding, ensuring stability and mitigating risks for the ambitious expansion.
As part of a broader collaboration, financial institutions from France, Britain, and Japan are also participating in the loans. Beyond enhancing Germany’s telecom security, the project is expected to benefit Japanese firms operating in the country by offering a trusted platform for handling sensitive data.
A decision to allocate satellite spectrum administratively rather than through an auction aims to increase competition in India’s vast telecom market. Telecoms Minister Jyotiraditya Scindia emphasised the government’s commitment to providing consumers with greater choice, despite concerns from Mukesh Ambani’s Reliance Jio over losing ground to Elon Musk’s Starlink. Reliance had pushed for auctions, arguing they ensure a level playing field after the company invested $19 billion in airwave rights.
Analysts suggest administrative allocation aligns with global norms and reduces investment barriers for foreign companies. Scindia noted that current satellite technology is limited to outdoor use, which distinguishes it from indoor services offered by terrestrial networks. Applications from Starlink and Amazon Kuiper to enter India’s satellite broadband market, projected to reach $1.9 billion by 2030, are under review.
India’s competitive telecom sector, with 942 million users and low data costs, is attracting significant global interest. Bharat Sanchar Nigam Limited (BSNL), a state-run operator with 99 million users, is expanding its 4G offerings to regain market share. Meanwhile, the government remains tight-lipped about plans to assist Vodafone Idea, which faces $24 billion in dues.
Musk’s disruptive approach, evident in markets like Kenya where Starlink’s pricing undercut local rivals, signals potential shifts in India’s broadband landscape. The new satellite policy could bring more innovation, fostering a dynamic environment for global and domestic players.
Donald Trump has rescinded a 2023 executive order issued by Joe Biden aimed at mitigating risks associated with AI to consumers, workers, and national security. Biden’s order mandated that developers of high-risk AI systems share safety test results with the US government before public release, under the Defense Production Act. It also required federal agencies to establish safety standards addressing potential threats such as cybersecurity, chemical, and biological risks. This move came amid congressional inaction on AI legislation.
The Republican Party had pledged to overturn Biden’s order, claiming it stifled AI innovation. The party’s 2024 platform emphasises support for AI development that aligns with free speech and human progress. Generative AI technologies, capable of creating content like text and images, have sparked both excitement and concern over their potential to disrupt industries and eliminate jobs.
While Trump revoked Biden’s AI safety framework, he left intact another executive order issued last week that supports the energy needs of advanced AI data centres. Biden’s newer order calls for federal assistance, including leasing Defense and Energy Department sites, to support the rapid growth of AI infrastructure. Meanwhile, US companies like Nvidia have criticised recent Commerce Department restrictions on AI chip exports, reflecting ongoing tensions between regulation and innovation in the tech sector.
Meta has announced a deal to purchase 200 megawatts of solar power from multinational utility Engie. The move bolsters the tech giant’s renewable energy portfolio, which now exceeds 12 gigawatts. The new solar farm, located in Texas, is near one of Meta’s existing data centres and is expected to become operational by 2025.
The push for renewable energy comes as tech companies face rising power demands driven by AI development and the rapid construction of data centres. Meta recently revealed plans for a 2-gigawatt data centre in Louisiana, relying on natural gas. The firm has also expressed interest in nuclear power, seeking proposals for up to 4 gigawatts of nuclear energy by the early 2030s.
While nuclear energy garners significant attention, renewable sources are crucial in powering today’s tech infrastructure. Meta’s solar energy deal mirrors efforts by other tech giants like Google and Microsoft, which have secured multi-billion-dollar renewable energy agreements. As companies race to meet energy needs, the speed of renewable energy deployment continues to offer a competitive edge over emerging nuclear options.
Progress on the EU Cybersecurity Certification Scheme (EUCS), stuck in a deadlock since 2019, remains uncertain as discussions are unlikely to advance in the first half of 2025. Despite efforts by Poland, which is leading the EU ministerial meetings until July, disagreements over sovereignty requirements continue to stall the process. The EUCS aims to help companies demonstrate that their ICT solutions meet cybersecurity standards for the EU market but has faced resistance, particularly from France, which wants to preserve its certification system, SecNum Cloud.
The European Cybersecurity Certification Group (ECCG) from ENISA has yet to provide an opinion on the scheme, with its next meeting possibly taking place in February. Poland plans to prioritise cybersecurity during its presidency, hosting key events like an informal telecom minister meeting in March and a conference on ENISA standardisation, though industry groups remain sceptical about a breakthrough.
Lobbyists, including the global software industry group BSA, have criticised the delays. They argue that cybersecurity standards should focus on technical protections rather than political considerations and have urged the Commission to adopt the scheme quickly to strengthen Europe’s cybersecurity resilience.
Further complicating matters, the EU Cybersecurity Act (CSA), which underpins ENISA’s authority to create certification schemes, is under evaluation but has not yet been revised. Of the three certification schemes proposed since 2019, only one has been adopted, with another for 5G still in progress. New EU Commissioner Henna Virkkunen has pledged to improve the adoption process for cybersecurity certification schemes as part of her mission to bolster Europe’s technological sovereignty and security.
Beijing-based AI company Zhipu Huazhang Technology has opposed the US government’s plan to add it to the export control entity list. The company argues the decision lacks a factual basis.
Zhipu issued a statement on its official WeChat account expressing strong opposition to the move. The firm criticised the US commerce department’s intentions, insisting the decision was unjustified.
Zhipu and its subsidiaries face restrictions on accessing US technologies if added to the list. The company maintains it operates lawfully and transparently in its business practices.
The US has been increasing scrutiny on Chinese technology firms, citing national security concerns. Zhipu emphasised its commitment to responsible technology development and cooperation with global partners.
Washington has blacklisted over two dozen Chinese entities, including Zhipu AI and Sophgo, for alleged links to restricted chip technology in Huawei processors. The Commerce Department has also tightened export controls on chips that could be diverted to Huawei.
Zhipu AI, backed by Tencent and Alibaba, was accused of contributing to China’s military modernisation through advanced AI research. Sophgo faced scrutiny after a chip found in Huawei’s Ascend 910B AI system matched one it ordered from Taiwan Semiconductor Manufacturing Co (TSMC).
The measures impose stricter licensing requirements for chip exports, targeting semiconductors at 14 or 16-nanometre nodes used in AI applications. New restrictions also affect DRAM memory, crucial for high-bandwidth AI processing, which could impact Chinese chipmaker CXMT.
Zhipu AI denied the claims, while Sophgo stated it had no direct or indirect ties to Huawei. Huawei and TSMC declined to comment on the latest sanctions, which build on previous curbs against Huawei and its network of suppliers.
